We present a probabilistic model that uses both prosodic and lexical cues for the automatic segmentation of speech into topically coherent units. We propose two methods for combining lexical and prosodic information using hiddenMarkov models and decision trees. Lexical information is obtained from a speech recognizer, and prosodic features are extracted automatically from speech waveforms. We e...

The selection of appropriate Lexical Units (LUs) is an important issue in the development of Continuous Speech Recognition (CSR) systems. Word has been used classically as unit in most of them. However, proposals of non-word units have begun to arise. Since the subject of this study is the Basque language, which is an agglutinative language with a complex structure inside words, non-word units ...

We present a reinterpretation of lexical information embedded in the English WordNet in an alternate type of structure called lexical system. First, we characterize lexical systems as graphs of lexical units (word senses) connected mainly by Meaning-Text lexical function relations, then introduce a hand-built lexical system: the French Lexical Network or frLN, a lexical resource that implements...

Speech recognition systems are conventionally broken up into phonemic acoustic models, pronouncing dictionaries in terms of the phonemic units in the acoustic model and language models in terms of lexical units from the pronouncing dictionary. Here we explore a new method for incorporating pronunciation probabilities into recognition systems by moving them from the pronouncing lexicon into the ...

We model word substitution errors made by normal and aphasic speakers with an interactive activation model of lexicalization. This comprises a three-layer architecture of semantic, lexical, and phonological units. We test four hypotheses about the origin of aphasic word substitutions: that they result from pathological decay, loss of within-level inhibitory connections, increased initial random...

We present an algorithm for local word grouping to extricate fixed word order dependencies in Hindi sentences. Local word grouping is achieved by defining regular expressions for the word groups. Ambiguities occurring during word grouping are also resolved. Hindi being a free order language, fixed order word group extraction is essential for decreasing the load on the free word order parser. Th...

Through ACT (Annotated Corpora of Text) software package for lexical and corpus processing of European written cultural sources (currently used for processing of mediaeval Slavonic manuscripts) this work presents another step forward towards a contextual and intelligent heritage Information Technology framework. ACT is suitable for capturing characteristics of old written sources including rich...

This work presents a software package ACT (Annotated Corpora of Text) for lexical and corpus processing of European written cultural sources (currently used for processing of mediaeval Slavonic manuscripts). I use ACT as a contribution towards a contextual and intelligent heritage Information Technology framework. The software is suitable for capturing characteristics of old written sources inc...

In this study, considering the effect of phrase grouping in spontaneous speech, prosodic words, instead of lexical words, are adopted as the units for error correction of speech recognition results. The prosodic words and the corresponding mis-recognized word fragments are obtained from a speech database to construct a mis-recognized word fragment table for the extracted prosodic words. For eac...

If the objective of a Continuous Automatic Speech Understanding system is not a speech-to-text translation, words are not strictly needed, and then the use of alternative lexical units (LUs) will bring us a new degree of freedom to improve the system performance. Consequently, we experimentally explore some methods to automatically extract a set of LUs from a Spanish training corpus and verify ...